16:00:12 <jungleboyj> #startmeeting cinder
16:00:13 <openstack> Meeting started Wed Apr 24 16:00:12 2019 UTC and is due to finish in 60 minutes.  The chair is jungleboyj. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:14 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:17 <openstack> The meeting name has been set to 'cinder'
16:00:22 <whoami-rajat> Hi
16:00:39 <jungleboyj> courtesy ping:  jungleboyj diablo_rojo, diablo_rojo_phon, rajinir tbarron xyang xyang1 e0ne gouthamr erlon tpsilva ganso patrickeast tommylikehu eharney geguileo smcginnis lhx_ lhx__ aspiers jgriffith moshele hwalsh felipemonteiro lpetrut lseki _alastor_ whoami-rajat yikun rosmaita enriquetaso hemna _hemna
16:00:49 <e0ne> hi
16:00:50 <smcginnis> o/
16:00:55 <jungleboyj> @!
16:00:55 <_alastor_> \o
16:00:55 <_pewp_> jungleboyj ( Β΄ β–½ ` )οΎ‰
16:00:57 <rajinir> o/
16:00:58 <rosmaita> o/
16:01:30 <xyang> hi
16:02:06 <jungleboyj> Come one, come all ... gather around the IRC hearth.
16:02:23 <jungleboyj> I will tell you stories of storage and summit planning.
16:02:34 <geguileo> hi! o/
16:03:17 <jungleboyj> Ok.  Guessing this is all we are getting for today.
16:03:31 <jungleboyj> #topic announcements
16:03:37 <jungleboyj> Have quite a few today.
16:03:47 <jungleboyj> First, we have the planning for the Train PTG.
16:04:02 <jungleboyj> #link https://etherpad.openstack.org/p/cinder-train-ptg-planning
16:04:04 <lseki> o/
16:04:05 <_erlon_> hey
16:04:15 <jungleboyj> As you can see I have started organizing the topics into time slots.
16:04:33 <jungleboyj> I think that lunch is going to be roughtly 12:30 to 1:30.  We can adjust if that is incorrect.
16:04:39 <jungleboyj> Lunch is provided at the PTG.
16:04:54 <jungleboyj> I have tried to put subject with whoami-rajat on Thursday morning.
16:05:04 <jungleboyj> Also am working around requests from the Glance team.
16:05:37 <jungleboyj> I will be finishing up the work in there today.  If you have concerns with the schedule please ping me on IRC or make a note in the etherpad.
16:05:43 * e0ne will miss some part of Thursday and Friday discussions :(
16:05:44 <whoami-rajat> jungleboyj: friday morning? and thanks!
16:06:07 <jungleboyj> whoami-rajat:  Sorry, yes, Friday morning.  Is that ok?
16:06:31 <whoami-rajat> jungleboyj: yes, early is good for me.
16:06:53 <jungleboyj> Ok.  I guess I could try to flip that back to Thursday morning so it isn't on your Friday night.
16:07:34 <jungleboyj> whoami-rajat:  Let me know your preference.
16:07:46 <whoami-rajat> jungleboyj: any morning is fine for me.
16:07:56 <jungleboyj> Ok.  I will leave things as is then.
16:08:09 <jungleboyj> We have our dinner plan:
16:08:20 <whoami-rajat> jungleboyj: ++
16:08:21 <jungleboyj> #link https://etherpad.openstack.org/p/cinder-train-dinner-plan
16:08:56 <jungleboyj> A few people from Glance may join us.
16:09:18 <whoami-rajat> jungleboyj: can we also assign the untyped vol discussion on thurs/friday morning?
16:09:34 <jungleboyj> I will make reservations today for ~15 so that we have room if others sign up.
16:09:40 <jungleboyj> whoami-rajat:  Oh, good point.  Yes.
16:09:41 <whoami-rajat> if that isn't much trouble.
16:10:07 <whoami-rajat> jungleboyj: thanks.
16:10:13 <jungleboyj> Welcome!
16:10:23 <jungleboyj> Thanks for joining remotely.
16:10:43 <whoami-rajat> jungleboyj: np :)
16:10:51 <jungleboyj> I have created the Forum Etherpads:
16:11:08 <jungleboyj> #link https://etherpad.openstack.org/p/denver-forum-cinder-direct-user-feedback
16:11:22 <jungleboyj> #link https://etherpad.openstack.org/p/denver-forum-cinder-improving-drvr-cap-rep
16:11:57 <jungleboyj> eharney:  If you can please put some notes as a starting point in the https://etherpad.openstack.org/p/denver-forum-cinder-improving-drvr-cap-rep
16:12:02 <jungleboyj> That would be appreciated.
16:12:13 <eharney> jungleboyj: yeah, i'm hoping to get a draft spec up with my thinking on that this week, but time is getting short
16:12:24 <jungleboyj> If everyone can help attend the User Feedback one to facilitate discussion it would be great.
16:13:49 <jungleboyj> eharney:  Ok, if you can't get a spec out there at least some notes for discussion in the etherpad would be appreciated.
16:13:58 <eharney> sure
16:14:55 <jungleboyj> eharney:  Thank you sir!
16:15:19 <jungleboyj> Appreciate anyone who can, joining those sessions.
16:15:37 <jungleboyj> I have had a Cinder Onboarding session added as one other group backed out.
16:15:48 <jungleboyj> Tuesday, May 30, 2019 - The Colorado Convention Center - 406 10:50 to 11:30
16:15:59 <jungleboyj> If you can join me to answer questions, it would be great.
16:16:20 <jungleboyj> Finally, Our project update:  Wednesday May 1, 2019 - The Colorado Convention Center - Meeting Room Level - 405  16:20 to 16:40
16:16:30 <jungleboyj> Come throw tomatoes and heckle if you wish.
16:16:38 <e0ne> :)
16:17:37 <jungleboyj> Ok.  So, that covers the announcements.
16:17:56 <jungleboyj> #topic Cross Project Discussions:
16:18:09 <jungleboyj> #link https://etherpad.openstack.org/p/ptg-train-xproj-nova-cinder
16:18:14 <jungleboyj> We have some time scheduled with Nova.
16:18:24 <jungleboyj> I believe there is just one topic in there.
16:18:48 <jungleboyj> Yeah.  Is anyone aware of other topics we need to touch on?
16:19:12 <smcginnis> Will be nice if we're down to just that.
16:19:22 <jungleboyj> smcginnis:  Indeed.
16:19:29 <jungleboyj> We have come a long way?
16:19:39 <jungleboyj> Lets rework the attachment flow again.
16:20:28 <jungleboyj> :-)  I am going to see if the Placement guys want to talk to us.
16:20:35 <jungleboyj> Otherwise I guess we are good there.
16:21:52 <jungleboyj> #topic releases update
16:21:56 <jungleboyj> rosmaita:
16:22:08 <rosmaita> #link https://etherpad.openstack.org/p/cinder-releases-tracking
16:22:18 <rosmaita> (at least they haven't change the etherpad url yet!)\
16:22:30 <rosmaita> all the April stable releases are set except for cinderclient
16:22:41 <rosmaita> i don't know if people have just been busy, or are having second thoughts about the backport strategy we discussed last week?
16:22:52 <rosmaita> to recap, the situation is that although the functionality supporting some API changes is in Rocky, the code change so that the client knows it can deal with the appropriate microversions wasn't made, so the stable/rocky client will only use 3.52 or less, although it can handle up to 3.55
16:23:05 <rosmaita> the patch to bring the client up to 3.55 has merged in master, it is proposed for stable/stein as
16:23:16 <rosmaita> #link https://review.opendev.org/#/c/654609/
16:23:31 <rosmaita> this would then need to be backported to stable/rocky, and then we will be ready to release from cinderclient from stable/rocky
16:23:42 <rosmaita> i just want to make sure everyone is onboard with that strategy before i make a pest of myself requesting reviews
16:23:59 <rosmaita> (pause for comments)
16:24:02 <jungleboyj> I think we just need to get another core to look at it and agree.  I am fine on the strategy.
16:24:08 <smcginnis> Speaking of backports, we should probably also talk about folks being onboard with upgrade checker backports.
16:24:09 <jungleboyj> Don't really have another option.
16:24:16 <jungleboyj> smcginnis:  ++
16:24:33 <rosmaita> ok, as far as stable/stein goes, we've got 2 things that have both merged to master
16:24:43 <rosmaita> #link https://review.opendev.org/#/c/653472/
16:24:52 <rosmaita> that one adds a tiny bit of functionality to support mv 3.56 (just prints the user_id in the backup-list response), and
16:25:03 <rosmaita> #link https://review.openstack.org/#/c/653473/
16:25:15 <rosmaita> #link https://review.openstack.org/#/c/653473/
16:25:18 <whoami-rajat> smcginnis:  is there any specific time before which we need to propose upgrade check?
16:25:31 <jungleboyj> whoami-rajat:  The sooner the better.
16:25:56 <rosmaita> that last one brings the supported mv to 3.58 (cinder stable/stein is at 3.59)
16:26:02 <smcginnis> whoami-rajat: We missed the cutoff for stein, so we have upgrade checks for stein in master. Those should probably be backported. Any new ones should get proposed to master as soon as they can be.
16:26:13 <rosmaita> i want to verify that we want to backport those to stable/stein before releasing an update of the stable/stein client?
16:26:28 <whoami-rajat> jungleboyj: smcginnis  ok, will propose the untyped volume one asap.
16:26:47 <jungleboyj> whoami-rajat:  Oh, that one will go with Train.  You don't need to hurry that.
16:27:07 <jungleboyj> That should be part of the untyped volume code you push up.  When you get there.
16:27:28 <whoami-rajat> jungleboyj: oh okay. **feww**
16:27:31 <jungleboyj> :-)
16:27:43 <jungleboyj> Ok, lets get back to rosmaita questions.
16:28:04 <jungleboyj> So, I think we want to merge the cinderclient MV changes/fixes.
16:28:18 <rosmaita> into stable/stein ?
16:28:24 <jungleboyj> And I think we want to backport them before doing a cinderclient release for stable/stein.
16:28:33 <jungleboyj> rosmaita:  Yes.
16:28:37 <rosmaita> ok, that's what i was looking for :)
16:28:47 <jungleboyj> smcginnis: eharney Any disagreement?
16:28:53 <smcginnis> ++
16:29:43 <eharney> i still wonder about what it means for anyone using get_highest_client_server_version
16:29:54 <eharney> but i kind of quit thinking about that after we talked about it before, so not sure..
16:30:04 <jungleboyj> rosmaita: I went and looked at all the other release proposals.
16:30:18 <rosmaita> jungleboyj: ty
16:30:34 <rosmaita> eharney: are you thinking a backward compat issue?
16:30:35 <jungleboyj> eharney:  I think it means they will get the right MV for Stein instead of the wrong one.
16:30:50 <eharney> rosmaita: yeah, i'm unclear on what we actually guarantee around that
16:31:58 <rosmaita> my impression is that it's queryable so that users don't hard code an mv?
16:32:18 <jungleboyj> rosmaita: ++
16:33:09 <jungleboyj> Ok.  Anything else on that?
16:33:35 <jungleboyj> Ok.
16:33:48 <jungleboyj> #topic periodic_interval config option
16:33:54 <jungleboyj> rosmaita:  You again.
16:33:58 <rosmaita> i was following up on a suggestion i made on a review that we should enhance the help text for our periodic_interval config option, but looking more closely,i think we are using it in too many places, sometimes at cross-purposes
16:34:05 <rosmaita> i put a bunch of details on the bug
16:34:11 <rosmaita> #link https://launchpad.net/bugs/1823748
16:34:13 <openstack> Launchpad bug 1823748 in Cinder "periodic_interval used in too many places" [Undecided,In progress] - Assigned to Brian Rosmaita (brian-rosmaita)
16:34:21 <rosmaita> and put up a patch
16:34:27 <rosmaita> #link https://review.opendev.org/#/c/651052/
16:34:35 <rosmaita> the reason i want to bring it up (besides being impatient), is that devstack has a CINDER_PERIODIC_INTERVAL setting that was introduced in havana
16:34:42 <rosmaita> the commit message said it was a workaround for a cinder bug, but when you look at the bug, it was fixed in a way that does not involve changing the periodic_interval
16:34:50 <rosmaita> so i think that this devstack thing is hanging around for no good reason and filed a bug
16:34:57 <e0ne> do we really need 4 new config options?
16:35:26 <rosmaita> e0ne: not sure, but they are doing 4 different things
16:35:59 <rosmaita> the devstack problem is that it shows up in a setup script that is used by *a lot* of projects
16:36:00 <jungleboyj> Hmmm, will need to look at that.
16:36:05 <eharney> they could default to $periodic_interval, perhaps
16:36:17 <rosmaita> yeah, maybe i should stop there
16:36:17 <e0ne> eharney: +1
16:36:32 <eharney> which supports the old behavior for people who don't want to worry about setting them all
16:36:38 <rosmaita> let's get the cinder side sorted out first and then i will worry about devstack
16:36:53 <eharney> but i do think it's right to use more than just one option
16:37:11 <jungleboyj> eharney:  That sounds reasonable.  Then we don't have to deprecate the old option as well.
16:37:22 <rosmaita> well, the old option would stick around
16:37:25 <smcginnis> In general, I think we've wanted to reduce the number of config options in Cinder. But if they do different things, then it makes sense for them to be different knobs.
16:37:31 <rosmaita> it would only be used for its specific purpose
16:37:43 <jungleboyj> Makes more tuning available if they are all different purposes.
16:37:54 <rosmaita> see what you think when you look at the bug, i have pointers to the code where they're used
16:37:58 <e0ne> smcginnis: that's exactly what I wanted to say
16:38:17 <jungleboyj> e0ne: ++
16:38:19 <smcginnis> eo;)
16:38:38 <smcginnis> Typing not working so much for me today. :D
16:38:39 <jungleboyj> Is that a train emoji?
16:39:35 <smcginnis> πŸš† πŸš„ πŸš… πŸš‚ 🚞 πŸšƒ
16:39:36 <whoami-rajat> seems like we can merge the backup config opts into one since they're not much different from default values or behavior.
16:40:11 <smcginnis> πŸ›€οΈ  πŸš‰
16:40:17 <smcginnis> OK, I'm done. :]
16:40:20 <jungleboyj> 🚌
16:40:45 * jungleboyj misses the dumpster fire
16:40:49 <rosmaita> ok, i'll look forward to comments on the review
16:40:55 <jungleboyj> rosmaita: ++
16:41:25 <jungleboyj> That is the end of the proposed agenda
16:41:28 <rosmaita> the devstack issue will be tricky because we might break the gate if that setting is actually making a difference
16:41:42 <rosmaita> that's all from me
16:41:46 <jungleboyj> #topic Open Discussion
16:41:53 <jungleboyj> rosmaita:  Is the man of the hour.
16:42:02 <rosmaita> :D
16:42:09 <jungleboyj> I have not heard any dissent on his core nomination so I am going to add him to the list.
16:42:24 <smcginnis> Woot! Welcome rosmaita
16:42:30 <jungleboyj> Tony also asked I make you stable core which I think is a great idea.  So, both of those will happen today.
16:42:36 <whoami-rajat> i would like some inputs on the patch https://review.opendev.org/#/c/641698/ . i still feel it's needed before/after cinderclient design discussion anyway.
16:42:49 <rosmaita> thank you, i will do my best to not destabilize cinder!
16:42:50 <e0ne> rosmaita: welcome to the team and thanks for all your contributions!
16:42:53 <jungleboyj> πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘
16:43:13 <whoami-rajat> rosmaita: congratulations and welcome!
16:43:29 <jungleboyj> rosmaita: It is well deserved.  Thank you for your help lately!
16:43:34 <e0ne> jungleboyj: +1 but only stable-core team can add rosmaita
16:43:59 <jungleboyj> e0ne:  Ah, that is right.
16:44:19 <smcginnis> Done
16:44:35 <jungleboyj> Wow, that was fast.  :-)
16:44:51 <smcginnis> Help when you know a guy. :D
16:45:00 * e0ne wonders if smcginnis is not core reviewer somewhere
16:45:00 <rosmaita> thanks!
16:45:14 <smcginnis> e0ne: Haha! :)
16:45:33 <smcginnis> e0ne: Not in horizon. ;)
16:45:37 <jungleboyj> smcginnis:  Is always the guy I go to when I need to know a guy.
16:45:47 <jungleboyj> Brian is now on the core list.
16:45:50 <e0ne> smcginnis: you just need few reviews... ;)
16:45:55 <jungleboyj> https://review.opendev.org/#/admin/groups/83,members
16:46:38 <jungleboyj> And on Stable Maint:  https://review.opendev.org/#/admin/groups/534,members
16:46:44 <jungleboyj> smcginnis:  Thanks for doing that so quick.
16:46:56 <smcginnis> Clickety click
16:47:17 <jungleboyj> So, that is all good stuff.
16:47:45 <jungleboyj> Anything else we need to discuss today?
16:48:06 <whoami-rajat> ^^ if my patch was missed in the celebration
16:48:12 <smcginnis> :)
16:48:58 <smcginnis> whoami-rajat: Is that in the PTG discussion etherpad?
16:49:05 <jungleboyj> πŸ‘πŸ‘πŸ‘
16:49:08 <jungleboyj> What?  ;-)
16:49:25 <whoami-rajat> smcginnis: dont think so.
16:49:26 <smcginnis> jungleboyj: whoami-rajat is actually trying to get some work done here. :)
16:49:33 <smcginnis> #link https://review.opendev.org/#/c/641698/
16:49:45 <jungleboyj> So, eharney is still -2 there.  So, think that needs to be addressed.
16:49:51 <jungleboyj> :-)
16:50:00 <eharney> oh yeah this thing
16:50:03 <rosmaita> i thought that was on the ptg agenda?
16:50:07 <whoami-rajat> jungleboyj:  already discussed
16:50:20 <jungleboyj> Ok.
16:51:40 <jungleboyj> eharney:  Are you removing your -2 based on this comments?
16:52:18 <eharney> aren't we going to figure out how this whole API should work at the PTG?
16:52:29 <whoami-rajat> its related and conflicts with our PTG discussions but this can be modified easily, since a lot of code needs to be modified after that
16:54:38 <whoami-rajat> eharney: yes. but it works correctly with the current flow. but i'm not sure so proposed in the meeting.
16:54:51 <whoami-rajat> not sure if we need to hold it*
16:55:02 <jungleboyj> Ok.  So, if it requires more discussion lets do so at the PTG.
16:55:20 <eharney> i don't think there really a concern about whether it worked correctly currently... it was more about how the whole user interaction of this is supposed to look
16:55:53 <rosmaita> that was my understanding too
16:56:18 <eharney> i think the current patch also does something a bit different from what the commit message describes, so i'm having trouble tracking where we are on this exactly
16:57:06 <jungleboyj> Ok.  So lets talk more next week.  Don't see a need to hurry this.
16:57:07 <whoami-rajat> eharney: oh yes, we moved it in the filter options then.
16:57:22 <whoami-rajat> okay if everyone agrees to discuss the same at PTG
16:57:34 <jungleboyj> ++
16:57:34 <rosmaita> works for me
16:57:58 <jungleboyj> Ok.  Anything else in the last 3 minutes.
16:58:43 <jungleboyj> Oh, just a note that we will not have a meeting next week due to the PTG.
16:58:57 <jungleboyj> Then we can return to our regularly scheduled time the following week.
16:59:24 <jungleboyj> Ok.  Thank you everyone for attending and look forward to seeing many of you next week.
16:59:30 <jungleboyj> #endmeeting